Apollo Graphql startup exceeded CPU time limit

I am trying to use graphql worker to serve as an Apollo Graphql server, using typescript / wrangler 2 / prisma edge.

All works fine, but it randomly fail (about 60% time) with wrangler complaining about:

notes: [
        text: 'Error: Script startup exceeded CPU time limit.\n [code: 10021]'

even though our worker is in usage_model = "unbound".